- From: Karim A. <directeur@gmail.com>
- Date: Wed, 21 Nov 2007 22:53:08 +0100
- To: "olivier Thereaux" <ot@w3.org>, bugs@timj.co.uk
- Cc: "W3C Validator Community" <www-validator@w3.org>
Hi Olivier, all, Thanks for the explanations. I actually don't agree with bug 4998. IMHO it's not a bug at all. Why? Because it's up to the server to decide for what it will be used. I mean, if the server is for production, well... let's cache and optimize. If, in the other hand, it's a dev one, it then could forget about caching and provide the validator with fresh content. Tim Jackson says: "The cache is not misbehaving by returning cached content" and it's true. Neither the validator is. Also true. The one thing that's somehow misbehaving is the server for not being clear about its intended use: dev or production? So it's not up to the validator to change its "normal" behavior as a "normal" client, but to the server to understand its goal. See what I mean? Karim -- http://xhtml-css.com Be Valid or die learning On 11/21/07, olivier Thereaux <ot@w3.org> wrote: > Hello Karim, all. > > On 15 nov. 07, at 23:47, Karim A. wrote: > > I read here: http://validator.w3.org/todo.html > > that in the 0.9.x series you'll start using Last-Modified > > to cache validation results and request again only > > if-modified-since. > > Yes, that would make things faster. For the moment, the caching > behavior of the validator is to send its requests with "Cache-control: > max-age=0" so that no cache between the validator and final server > decided to send a cached and outdated version. > > http://www.w3.org/Bugs/Public/show_bug.cgi?id=4998 > > Having the validator use a cache would be nice, but complicated to > implement. It's there in the todo, but some items there may or may not > be implemented in the end. > > -- > > olivier >
Received on Wednesday, 21 November 2007 21:53:15 UTC